This is a revised application for 3 years of funding for two sites and 5 consortium sites (PA 05-106, "Deep Sequencing and Haplotype Profiling of Mental Disorders"). The goal is to identify and characterize genetic variation that contributes to schizophrenia (SZ) susceptibility, by carrying out a genome-wide association GWA) study followed by resequencing, genotyping and biological experiments. Two samples will be studied: 3,000 SZ and 3,000 control subjects of European ancestry (EA), and 1,200 cases and 1,200 controls of African-American (AA) ancestry. The GWA datasets will include 550,000 SNPs in the EA sample (the revised Affymetrix 500K array and 50K Gene-Focused chip that includes 20K nsSNPs), and the new Affymetrix 1M array in the AA sample (the 500K array and 500K additional SNPs with increased coverage of African variation). The new 500K array also provides genomewide assays of additional copy number variants (CNVs). The Genetic Association Information Network (GAIN) will genotype 1450/1450 EA cases/controls (500K) and the entire AA sample (1M). The Affymetrix consortium site will genotype the remaining 1550/1550 EA cases/controls with the 500K array, and all EA subjects with the 50K chip. Preliminary statistical studies are proposed, to select an optimal data analysis strategy that tests every HapMap SNP using single- and multi-marker tests, evaluates evidence for association on European- and African-ancestry chromosomes (after inferring local ancestry in admixed individuals) and in the combined data, controls for subtle population substructure, and evaluates empirical p-values through permutation. A set of" 15 candidate intervals will be selected based on p-value threshold, Rank Truncated Product analysis, replication experiments, and bioinformatic and biological information. Deep resequencing experiments will detect any significant case-control difference in rare functional mutations, and will discover new rare and common SNPs. Further genotyping of each region will include rare/functional SNPs and additional common SNPs for optimal tagging of common variants. Based on evidence for association and available information about each gene, the associated genomic interval, and the associated variants, biological studies will be undertaken to begin to evaluate the functional effects of these variants and the implications for hypothesis about mechanisms underlying susceptibility to SZ. [unreadable] [unreadable] [unreadable]